Image index model for retrieval using Hausdorff distortion

An image index model has been proposed that uses vector quantization (VQ) to index the images. This index structure uses the Hausdorff distortion (HD) to enable faster retrieval of images from an image database. The authors have compared the HD value with the Modified Hausdorff distance (MHD) and Encoding Distortion Distance (EDD). The results show higher precision for the HD values whereas the MHD values give lesser precision for the same recall. This reduces the processing needed to construct the index.